Bangladesh

# packages

library(wdpar)
library(sf)
library(dplyr)
library(leaflet)
library(ggplot2)
library(stringr)
library(plotly)
library(reactable)
library(reactablefmtr)
library(showtext)

## settings
# set country
# 'Bangladesh' <- 'Ecuador'

# set colours
background_colour <- '#faf7f2'

# set fonts
font_add_google("Work Sans", family = "worksans")
showtext_auto()


# read base data
gadm <- geodata::gadm(
  country = 'Bangladesh',
  path = ".",
  level = 0,
  resolution = 1 # high
) |>
  st_as_sf() |>
  janitor::clean_names()

raw_pa_data <- wdpa_fetch(
  x = 'Bangladesh',
  wait = FALSE,
  download_dir = "../wdpar"
) |>
  janitor::clean_names()


# raw_pa_data |>
#   st_drop_geometry() |>
#   count(realm)
#
# raw_pa_data |>
#   filter(site_id == 198315) |>
#   glimpse()

basemap_leaflet <- leaflet() |>
  addMapPane("background", zIndex = 0) |> # Level 1: bottom
  addMapPane("rasters", zIndex = 1) |> # Level 2: middle
  addMapPane("polygons", zIndex = 200) |> # Level 3: middle
  addMapPane("polylines", zIndex = 300) |> # Level 3: middle
  # addMapPane("rasters", zIndex = 100000) |>    # Level 3: middle
  addMapPane("points", zIndex = 440) |> # Level 4: middle
  addMapPane("labels", zIndex = 450) |> # Level 5: top

  addProviderTiles(
    providers$Esri.WorldStreetMap,
    group = "Open Street Map",
    options = pathOptions(pane = "background")
  ) |>
  addProviderTiles(
    providers$Esri.WorldImagery,
    group = "Satellite",
    options = pathOptions(pane = "background")
  ) |>
  addPolygons(
    data = gadm,
    weight = 1.5,
    fillOpacity = 0,
    color = '#333333',
    opacity = 1,
    group = 'Country boundary',
    label = ~ htmltools::htmlEscape(country),
    options = pathOptions(pane = "polygons")
  ) 

Key Biodiversity Areas (KBAs)

first stats

file_list <- list.files(
  '../../base_gis/KBAsGlobal_2025_September_02',
  pattern = '\\.shp$',
  full.names = T
)

df_list <- list.files(
  '../../base_gis/KBAsGlobal_2025_September_02',
  pattern = '\\.xlsx$',
  full.names = T
)
# readxl::excel_sheets(df_list)

kbas_0 <- lapply(file_list, read_sf) |>
  bind_rows() |>
  janitor::clean_names()


kbas <- kbas_0 |>
  filter(country == 'Bangladesh')


df <- readxl::read_xlsx(df_list) |>
  janitor::clean_names() |>
  filter(sit_rec_id %in% kbas$sit_rec_id)


kbas_1 <- kbas |>
  left_join(df) |>
  select(
    sit_rec_id,
    nat_name,
    int_name,
    kba_status,
    ends_with('status'),
    class,
    taxon_id,
    common_name,
    scientific_name,
    kba_criterias,
    year
  )

kbas_2 <- kbas_1 |>
  mutate(
    class = str_to_title(class),
    triggered_by_class = if_else(!is.na(kba_criterias), class, 'none'),
    triggered_by_scinam = if_else(
      !is.na(kba_criterias),
      scientific_name,
      'none'
    )
  )

kba_3 <- kbas_2 |>
  distinct(sit_rec_id, class, triggered_by_class, year) |>
  arrange(class) |>
  group_by(sit_rec_id) |>
  summarise(
    class = paste0(class, collapse = ', '),
    triggered_by = paste0(triggered_by_class, collapse = ', '),
    triggered_by_birds = if_else(str_detect(triggered_by, 'Aves'), 1, 0),
    year = unique(year)
  ) |>
  mutate(
    contain = case_when(
      class == 'Aves' ~ 'Only birds',
      str_detect(class, 'Aves') ~ 'Birds and others',
      !str_detect(class, 'Aves') ~ 'No birds',

      TRUE ~ 'check'
    )
  )


kba_3$triggered_by <- str_remove_all(kba_3$triggered_by, 'none|, ')

# kba_3 |>
#   count(contain, triggered_by_birds)

kbas_leaflet <- kbas_2 |>
  mutate(
    sci_name_class = paste0(scientific_name, ' (', class, ')'),
    comm_name_class = paste0(common_name, ' (', class, ')')
  ) |>
  arrange(class) |>
  # select(class, scientific_name, sci_name_class) |>
  group_by(sit_rec_id, int_name, nat_name, kba_status, year) |>
  summarise(
    scientific_name = paste0(sci_name_class, collapse = ', '),
    common_name = paste0(comm_name_class, collapse = ', ')
  ) |>
  left_join(kba_3) |>
  ungroup() %>%
  filter(st_geometry_type(.) != 'POINT') |>
  st_cast('MULTIPOLYGON')


kbas_leaflet_c <- kbas_leaflet |>
  st_drop_geometry() |>
  count(contain, triggered_by_birds, sort = T) |>
  mutate(contain = forcats::as_factor(contain), n = as.integer(n))

n_kbas <- length(unique(kbas_leaflet$sit_rec_id))


n_kba_birds <- kbas_leaflet |>
  st_drop_geometry() |>
  filter(contain != 'No birds') |>
  nrow()


# kbas_leaflet_c |>
#   ggplot() +
#   aes(y = contain, x = n, fill = n) +
#   geom_col() +
#   theme_minimal()
kbas_leaflet_c |>
  mutate(
    triggered_by_birds = if_else(triggered_by_birds == '1', 'Yes', 'No'),
    triggered_by_birds = factor(triggered_by_birds, levels = c('Yes', 'No'))
  ) |>
  arrange(triggered_by_birds, contain) 

# A tibble: 3 × 3
  contain    triggered_by_birds     n
  <fct>      <fct>              <int>
1 Only birds Yes                    1
2 Only birds No                    12
3 No birds   No                     7

  # ggplot() +
  # aes(x = contain, y = n, fill = triggered_by_birds) +
  # geom_col() +
  # labs(x = NULL, y = 'Count') +

  # scale_fill_manual('Triggered by birds?', values = c('#333333', '#d6d4cfff')) +
  # # scale_fill_manual('Triggered by birds?', values = c('#7D804BFF', '#E6EEE0FF')) +
  # theme_minimal() +
  # theme(
  #   legend.position = 'bottom',
  #   plot.background = element_rect(fill = background_colour),
  #   axis.title = element_text(family = 'worksans')
  # )

Total number of KBAs

20

Number of KBAs with Birds

13

# glimpse(raw_pa_data)
# st_geometry_type(raw_pa_data) == 'MULTIPOLYGON'
raw_pa_data_pols <- raw_pa_data |>
  filter(st_geometry_type(raw_pa_data) == 'MULTIPOLYGON') |>
  mutate(
    iucn_cat = factor(
      iucn_cat,
      levels = c(
        'Ia',
        'Ib',
        'II',
        'III',
        'IV',
        'V',
        'VI',
        'Not Applicable',
        'Not Reported',
        'Not Assigned'
      )
    ),
    realm = factor(realm, levels = c('Coastal', 'Marine', 'Terrestrial'))
  )

# raw_pa_data_pols$iucn_cat
# raw_pa_data |>
#   st_drop_geometry() |>
#   count(status)

KBAs map

# kbas_leaflet$contain |> unique() |> sort()

kbas_leaflet <- kbas_leaflet |>
  arrange(contain) |>
  mutate(contain = forcats::as_factor(contain))


pal_kbas <- colorFactor(
  domain = levels(kbas_leaflet$contain),
  palette = c('#2e6ba6', '#e53d3d', '#59de9e')
)

iucn_palette <- tribble(
  ~id, ~cat, ~hex_colour,
  'Ia', 'Ia Strict nature reserve', '#174721', 
  'Ib', 'Ib Wilderness area', '#7AC141',
  'II', 'II National park', '#5D873B',
  'III', 'III Natural monument or feature', '#AFBD36',
  'IV', 'IV Habitat/species management area', '#C8DF8D',
  'V', 'V Protected landscape/seascape', '#15477C',
  'VI', 'VI Protected area with sustainable use of natural resource', '#7EA1B5',
  'Not Applicable', 'NA Not Applicable', '#894444',
  'Not Assigned', 'NAS Not Assigned', '#E60000',
  'Not Reported', 'NR Not Reported', '#898944'
)

iucn_palette <- iucn_palette |> 
  filter(id %in% levels(raw_pa_data_pols$iucn_cat))

pal <- colorFactor(
  domain = levels(raw_pa_data_pols$iucn_cat),
  palette = iucn_palette$hex_colour
)
# pal <- colorFactor(
#   domain = levels(raw_pa_data_pols$iucn_cat),
#   c(
#     '#232021',
#     '#426f41',
#     '#59ba49',
#     '#c1178d',
#     '#feb016',
#     '#e61e24'
#   )
# )

basemap_leaflet |>
  addPolygons(
    data = kbas_leaflet,
    fillOpacity = 1,
    fillColor = ~ pal_kbas(contain),
    stroke = TRUE,
    color = '#333333',
    weight = ~ (triggered_by_birds * 2) + 0.5,
    opacity = 1,
    group = 'KBAs',
    # label = ~ htmltools::htmlEscape(desig_eng),
    options = pathOptions(pane = "polygons"),
    popup = paste0(
      '<b>',
      kbas_leaflet$int_name,
      '</b>',
      '</br>',
      '<a href="https://www.keybiodiversityareas.org/site/factsheet/',
      kbas_leaflet$sit_rec_id,
      '/assessment", target="_blank">Factsheet</a>',
      '</br>',
      '</br>',
      '<b>Groups: </b>',
      kbas_leaflet$contain,
      '</br>',
      '<b>National site name: </b>',
      kbas_leaflet$nat_name,
      '</br>',
      '<b>Triggered by: </b>',
      kbas_leaflet$triggered_by,
      '</br>',
      '<b>Year of last assessment: </b>',
      kbas_leaflet$year
    )
  ) |>

  addPolygons(
    data = raw_pa_data_pols,
    weight = 0.5,
    fillOpacity = 0.7,
    fillColor = ~ pal(iucn_cat),
    # stroke = FALSE,
    color = '#333333',
    opacity = 0.7,
    group = 'Protected areas',
    label = ~ htmltools::htmlEscape(desig_eng),
    # label = ~ htmltools::htmlEscape(iucn_cat),
    options = pathOptions(pane = "polygons"),
    popup = paste0(
      '<b>',
      raw_pa_data_pols$name_eng,
      '</b>',
      '</br>',
      '</br>',
      '<b>Designation: </b>',
      raw_pa_data_pols$desig_eng,
      '</br>',
      '<b>IUCN category: </b>',
      raw_pa_data_pols$iucn_cat,
      '</br>',
      '<b>Realm: </b>',
      raw_pa_data_pols$realm,
      '</br>',
      '<b>Status (year): </b>',
      raw_pa_data_pols$status,
      ' (',
      raw_pa_data_pols$status_yr,
      ')'
    )
  ) |>
  addLegend(
    data = kbas_leaflet,
    pal = pal_kbas,
    values = levels(kbas_leaflet$contain),
    opacity = 0.9,
    group = 'KBAs',
    title = 'KBAs'
  ) |>
  addLegend(
    data = raw_pa_data_pols,
    pal = pal,
    values = levels(raw_pa_data_pols$iucn_cat),
    opacity = 0.9,
    group = 'Protected areas',
    title = "Protected areas </br> IUCN cats"
  ) |>
  addLayersControl(
    baseGroups = c(
      # "Topography",
      "Open Street Map",
      "Satellite"
      
      # "Terrain",
    ),
    overlayGroups = c(
      'KBAs',
      'Protected areas'
    ),
    options = layersControlOptions(collapsed = FALSE, position = "bottomright")
  ) |>
  hideGroup(c('Protected areas')) |>
  addScaleBar(position = c("bottomleft"), options = scaleBarOptions()) 

Protected Areas (PAs)

first stats

raw_pa_data_pols |>
  st_drop_geometry() |> 
  count(desig_eng) |> 
  filter(str_detect(desig_eng, 'UNESCO-MAB'))

# A tibble: 0 × 2
# ℹ 2 variables: desig_eng <chr>, n <int>

g_labels <- levels(raw_pa_data_pols$iucn_cat)[
  length(levels(raw_pa_data_pols$iucn_cat)):1
]
n_desig <- raw_pa_data_pols |>
  filter(!str_detect(desig_eng, 'UNESCO-MAB')) |>
  pull(desig_eng) |>
  unique() |>
  length()
pas_count <- raw_pa_data_pols |>
  filter(!str_detect(desig_eng, 'UNESCO-MAB')) |>
  st_drop_geometry() |>
  count(iucn_cat, desig_eng, realm) |>
  mutate(iucn_cat = factor(iucn_cat, levels = g_labels))

palette_desig <- paletteer::paletteer_c("scico::roma", n = n_desig)
# palette_desig <- paletteer::paletteer_c("viridis::viridis", n = n_desig)
# palette_desig <- paletteer::paletteer_c("scico::batlow", n = n_desig)

ggplotly(
  pas_count |>
    arrange(iucn_cat) |>
    ggplot() +
    aes(y = iucn_cat, x = n, fill = desig_eng) +
    geom_col() +
    labs(y = 'IUCN categories', x = 'Count') +
    scale_fill_manual(
      values = palette_desig,
      # labels = \(x) stringr::str_wrap(x, width = 32),
      # 'Designation'
    ) +
    theme_minimal() +
    theme(
      legend.position = 'none',
      plot.background = element_rect(fill = background_colour),
      axis.title = element_text(family = 'worksans')
    ) +
    facet_wrap(~realm, scales = 'free')
) 

select(iucn_palette, -hex_colour)

# A tibble: 10 × 2
   id             cat                                                       
   <chr>          <chr>                                                     
 1 Ia             Ia Strict nature reserve                                  
 2 Ib             Ib Wilderness area                                        
 3 II             II National park                                          
 4 III            III Natural monument or feature                           
 5 IV             IV Habitat/species management area                        
 6 V              V Protected landscape/seascape                            
 7 VI             VI Protected area with sustainable use of natural resource
 8 Not Applicable NA Not Applicable                                         
 9 Not Assigned   NAS Not Assigned                                          
10 Not Reported   NR Not Reported                                           

colours_desig <- tibble(palette_desig, desig = sort(unique(pas_count$desig_eng)))
prep <- colours_desig |>
  select(colour = palette_desig, desig_eng = desig) |> 
  right_join(pas_count)

prep |> 
  reactable(
    showSortable = T,
    defaultSorted = list(n = 'desc'),
    searchable = TRUE,
    defaultPageSize = 20,
    showPageSizeOptions = TRUE,
    pageSizeOptions = c(10, 20, 50, 100, nrow(pas_count)),
    fullWidth = F,
    columns = list(
      colour = colDef(
        sortable = F,
        name = '',
        maxWidth = 100,
        style = function(value) {
          colour <- value
          list(background = colour, color = colour)
        }
      ),
      desig_eng = colDef(
        name = 'Designation',
        minWidth = 400
      ),
      realm = colDef(
        name = 'Realm'
      ),
      iucn_cat = colDef(
        name = 'IUCN (cat)',
        minWidth = 150
      ),
      n = colDef(
        name = 'Count'
      )
    )
  )